NYU CUSSL

E5 series (e.g., `intfloat/e5-large-v2`)

Strong, open-source embeddings for semantic search.

Excellent performance on MTEB benchmarkOpen-source and accessibleGood for retrieval tasksVarious model sizes available

Where it ranks today

Best for / Not great for

Best for
  • Semantic search engines
  • RAG implementation
  • Document retrieval
  • Academic and research applications
Not great for
  • Highly specialized classification tasks without fine-tuning
  • Users preferring fully managed, high-touch enterprise solutions
  • Real-time conversational AI requiring very fast re-ranking

Why it ranks here

The E5 model series remains a benchmark for open-source embeddings, consistently achieving top scores on retrieval benchmarks like MTEB. Its accessibility and strong performance make it a go-to for many RAG and semantic search projects.

30-day trend

Score breakdown

Search trends90
Benchmarks91
Developer buzz89
News mentions85

Pricing

API: $0.00 in · $0.00 out per 1M tokens · Consumer: $0.00/mo

Pricing plans

Popular
Open Source
Deploy E5 models yourself.
Free
  • Downloadable weights
  • Multiple model sizes
  • Full customizability
  • Requires self-hosting
Download E5 models
API Access (Example)
Use E5 via managed APIs.
$0 /usage
  • Pay-per-token
  • Managed service
  • Simple integration
  • Scalable inference
Use inference API
Compare with another modelHow is this score calculated? →Snapshot 2026-05-19